Previous research has shown that cognitive psychotherapy (CT) reduces or eliminates symptoms in many outpatients with nonpsychotic, nonbipolar major depression. One followup study suggests that prophylaxis ensues in 56% of those treated with CT in the year following treatment. However, not all patients respond. Roughly 60% attain a complete clinical remission by end of treatment. No consistent predictors of acute or prophylactic response to CT are known. Eighty male and female outpatients (aged 21-60 years) with nonpsychotic, nonbipolar major depression will be evaluated by SADS-L interviews, Resarch Diagnostic Criteria, symptom severity, history of illness, and selected cognitive measures, as well as by two consecutive nights of sleep EEG prior to treatment with 10 weeks (20 sessions) of cognitive therapy. Two groups (forty each) will be formed based on pretreatment REM latency (RLs) values. The main predictions to be tested are: those with low RLs (less than 60 minutes) will respond less well acutely and prophylactically than will those with RLs of 60 minutes or greater. Whether the acute or prophylactic response to CT can be predicted by history of illness (e.g., length of episode, length of illness, etc.) or cognitive measures (e.g., dysfunctional attitudes, self-control scale sources, etc.) will also be evaluated. Finally, whether symptomatic response is associated with cognitive or sleep EEG changes will be assessed.